sures and inducements would turn out to be less painful than decline under a process of later "natural" demographic adjustment, typically elicited by acute economic distress. Such pressures and inducements should not aim at changing the objective economic conditions of families in order to discourage fertility. To the contrary, the pure effects of income change, hopefully, will be consistently positive as development proceeds and as the economic status of the family improves. The policy should focus on changing the subjective image of these conditions in peoples' minds by exposing them to new knowledge and by manipulating their tastes, expectations, ambitions, and time horizons. An elaboration on this theme however is outside the limits of this paper. The foregoing considerations would powerfully reinforce the argument concerning the economic usefulness of family planning programs, already well established even in the absence of externalities, and suggest that the total program size should be geared to the level of the demand that is being generated, or be determined by organizational and technical rather than budgetary constraints. As to the distribution of funds within the total budget, as long as a sizable demand for the services offered remains to be satisfied, the principle of equating the marginal productivity of the funds in the various subprograms should be applied. This will require continued collection and analysis of program data as to cost-effectiveness, as well as data gauging attitudinal changes and measuring the potential demand for various program outputs.* As a general rule, however, great caution should be exercised not to promote particular types of programs in preference to others on the basis of observed or anticipated short-term results. The returns to programs are necessarily lagged, and the lags are likely to differ significantly from program to program. Informed wisdom under such conditions will be a better guide for action than a narrowly conceived, cost-effectiveness analysis. It is equally important that policy decisions take into account not only the direct demographic effects, but the sociopsychological context of the program and its far-reaching economic ramifications as well. REFERENCES 1. Coale, Ansley J., and Edgar M. Hoover, Population Growth and Economic Development in Low-Income Countries. Princeton: Princeton Univ. Press, 1958. *For an analysis of the richest body of data to date see (30). A cost-effectiveness study of selected family planning programs is currently under way under the direction of Professor Warren C. Robinson and will be made public later this year.